A mining algorithm based on time series association rules
Liping Liu, Ninghai Cui · 2012
Based on the concept lattice theory, this paper studies the cycle of association rule mining about the periodic fluctuations of time-series. First of all, do anti-season pretreatment to time-series, then give the algorithm of the generation of the cycle of association rules, and the pruning to the generated concept within algorithm improve the efficiency of the mining speed. And then we can use the given higher precision model to do anti-season changes calculation to those time series which dissatisfy anti-season pretreatment using the moving average method.